3/15/2006

Does the Superbowl Predict Economic Growth?

In one of my economics classes I remember hearing about something called the Super Bowl effect. For those of you who don't know what this is, I will explain. The myth says that if a team from the AFC wins the Superbowl, the national economy will do bad. If a team from the NFC wins the Superbowl, the national economy will do great. By just looking at the raw data it seams that this could be true. During the 1970's and early 1980's the AFC was strong and the economy was weak. The economy was strong during the rest of the 1980's when the NFC was strong. Mike Moffatt did a statistical study to determine if this myth had any truth. He looked at the last 39 Superbowls. He found that there is a statistical correlation between Superbowl result and economic growth. He found that the correlation between NFC wins and economic growth was .074963. I thought this was very interesting. Could the Superbowl really be correlated to the economy? If this is true it just goes to prove how complex the economy really is. It would be interesting to do similar studies and see what you would find. Does anyone have any ideas about why these two things would be correlated?

2 comments:

rico said...

The Super Bowl spurs economic growth in beer and chips. That's all that I have heard.

Dr. Tufte said...

The trick here is the word correlation. That doesn't mean causation. This leaves us open to statements like a good economy causes the NFC team to win.

Those who have taken me for other classes know my story that the arrival of storks along the southeastern coast of the North Sea is in fact correlated with births of babies.

For what it is worth, I did duplicate Moffat's regression. What he doesn't tell you is that the coefficient isn't significant. Another way of looking at that is that the confidence interval on the coefficient goes from .044 to -.024. This converts to a Steelers victory by 11 points converting into a change in growth from 3% to somewhere between 2.5% and 3.3%. The problem with that from a statistical perspective is that no change at all is pretty much right in the middle.

Sorry to rain on your parade Ashton. On the other hand, there is also a negative correlation between the length of women's skirts and the NFC margin of victory (i.e., NFC wins are associated with shorter lengths).